Adaptive Evolution of the Spike Protein in Coronaviruses

Abstract

Coronaviruses are single-stranded, positive-sense RNA viruses that can infect many mammal and avian species. The Spike (S) protein of coronaviruses binds to a receptor on the host cell surface to promote viral entry. The interactions between the S proteins of coronaviruses and receptors of host cells are extraordinarily complex, with coronaviruses from different genera being able to recognize the same receptor and coronaviruses from the same genus able to bind distinct receptors. As the COVID-19 pandemic has developed, many changes in the S protein have been under positive selection by altering the receptor binding affinity, reducing antibody neutralization activities, or affecting T-cell responses. It is intriguing to determine whether the selection pressure on the S gene differs between SARS-CoV-2 and other coronaviruses due to the host shift from nonhuman animals to humans. Here, we show that the S gene, particularly the S1 region, has experienced positive selection in both SARS-CoV-2 and other coronaviruses. While the S1-NTD domain exhibits signals of positive selection in the pairwise comparisons in all four coronavirus genera, positive selection is primarily detected in the S1-CTD domain (the receptor-binding domain, RBD) in the ongoing evolution of SARS-CoV-2, possibly owing to the change in host settings and the widespread natural infection and SARS-CoV-2 vaccination in humans.

Natural selection shapes the evolution of SARS-CoV-2 Omicron in Bangladesh

“…CoV-2 (Tang et al, 2023). In this study, we verified the spread and strength of natural selection acting on SARS-CoV-2 Omicron genomes circulating in Bangladesh.…”

“…A recent publication provided evidence of strong positive selection acting on the S gene during the evolution of SARS-CoV-2 ( Tang et al, 2023 ). In this study, we verified the spread and strength of natural selection acting on SARS-CoV-2 Omicron genomes circulating in Bangladesh.…”

Alterations of SARS-CoV-2 Evolutionary Dynamics by Pharmaceutical Factors

“…[39] Despite apparent advantages in Israel's favor, Israel still suffered a higher per-capita COVID-19 death rate than that for Palestine. [40] Vaccination alone cannot stop the spread of new variants, [41] and evidence of selective pressures on the spike protein, [2,[42][43][44][45] the target antigen of vaccine-generated immune responses, has also been found. [46] Mechanistically, given that most vaccines solely encode the spike protein, mutations that alter the conformation of the spike protein may prevent antibodies from binding to earlier variants of spike protein [Figure 1].…”

Multiple Lines of Evidence Support 199 SARS-CoV-2 Positively Selected Amino Acid Sites

“…Even when looking at PSSs identified by all three methodologies in the S protein (173, Table 1, PSS-3), the number of PSSs is high when comparing with the recently published literature. For instance, Tang et al [69], using codeML models M7 and M8 (accepting only those sites with a BEB score larger than 0.95 as a PSS) and 50 randomly chosen SARS-CoV-2 genomes, identified only 12 PSSs at the S protein. The difference between the two numbers is partially explained by the size of the sequence dataset, because the 50 randomly chosen genome sequences used by Tang et al [69] only represent a fraction of all the amino acid polymorphisms assayed in the COV2Var database.…”

“…For instance, Tang et al [69], using codeML models M7 and M8 (accepting only those sites with a BEB score larger than 0.95 as a PSS) and 50 randomly chosen SARS-CoV-2 genomes, identified only 12 PSSs at the S protein. The difference between the two numbers is partially explained by the size of the sequence dataset, because the 50 randomly chosen genome sequences used by Tang et al [69] only represent a fraction of all the amino acid polymorphisms assayed in the COV2Var database. Nevertheless, even when comparing with the 52 PSSs identified by us, using multiple medium-sized datasets, where all the years from 2019 to 2023 are equally represented to avoid a bias toward the years where the sequencing effort was heavier, the two numbers are still very different (173 vs. 52).…”